Contextual patch-NetVLAD: Context aware patch feature descriptor and patch matching mechanism for visual place recognition

This is the UTown7 dataset used in our paper, entitled "contextual patch-NetVLAD: Context aware patch feature descriptor and patch matching mechanism for visual place recognition".

It consists of images that are gathered from the Stephen Riady Centre, located on the campus of the National University of Singapore. They are gathered at seven unique locations across three different levels of the centre, including flavour back door, flavour front door, gym, swimming pool entrance, bank, restroom, and convenience store. It also contains the code to apply image augmentation methods to generate more images for the experimental performance evaluation.